Intergovernmental transfers in brazil : corruption and deforestation

Abstract

This paper investigates the effect of additional governmental transfers on several different measures of deforestation for the years between 1997 and 2016. We use comprehensive datasets from the MapBiomas database for our deforestation calculations. This relatively new database affords us the opportunity to analyze changes in Brazilian municipalities from a single data source over an extended period rather than having to draw from separate sources with different methodologies. We use a regression discontinuity design to estimate the causal effect of increases in transfers on deforestation in Brazil, emphasizing on municipalities in the Legal Amazon. Results obtained from the entire sample show that larger transfers result in significant increases in gross deforestation. Extending our analysis to pooled thresholds and pooled years, we find some evidence of separate patterns between the period from 1997–2006 and the period from 2008–2016, but mixed effects in both magnitude and statistical significance when we restrict our samples to individual thresholds. While there is evidence in the first period that is in line with our expectation of larger transfers leading to higher levels of deforestation, the surprising positive increases in forest coverage in the second period 2008–2016 suggest the promising possibility of governmental transfers having beneficial impacts on efforts to reduce deforestation in Brazil as well.